Posts by budgibson@me.dm
 (DIR) Post #ATmk2AD3sdNaatqLSK by budgibson@me.dm
       2023-03-19T20:52:20Z
       
       0 likes, 0 repeats
       
       @gpshead @simon @simon I think more generally, we respond to things we perceive as human as though they were full functioning humans. Witness how we interact with babies, the infirm, those on their deathbeds, those in catatonic states. Part of the ethics here has to be that the #LLM fully disabuses people of this tendency. I think GitHub CoPilot is a good example of an #LLM that succeeds on that score and is quite useful.
       
 (DIR) Post #ATmmN2IcgYm83BWc8e by budgibson@me.dm
       2023-03-19T21:17:23Z
       
       0 likes, 0 repeats
       
       @simon @gpshead I’m a fan of her reasoning on this score. It’s a sort of insight that drives readily from sociolinguistics. A large proportion of our utterances are phatic.
       
 (DIR) Post #ATtYIgJ6TIUhQsUyPI by budgibson@me.dm
       2023-03-23T03:47:05Z
       
       0 likes, 0 repeats
       
       @grammargirl I do think that learning more than one language changes how you think. It can be like living in two slightly different worlds at once.  It’s interesting that different languages are stored in different parts of the brain depending somewhat on the order you learned them.
       
 (DIR) Post #ATzcFiT0PnuD8oBTlY by budgibson@me.dm
       2023-03-26T01:59:39Z
       
       0 likes, 0 repeats
       
       @grammargirl @Jantar I think there are two views rooted in the philosophy of meaning: one is that the meaning of the whole is derived from the component parts (words) and the other that the component parts derive their meaning from the greater whole. My personal experience is that people from different language communities live in different zeitgeists and evolve their communications to suit. It’s a complicated topic with a lot of back and forth relationships.
       
 (DIR) Post #ATzdnZEdLSdUJEAPxY by budgibson@me.dm
       2023-03-26T02:16:58Z
       
       0 likes, 0 repeats
       
       @grammargirl @Jantar Here’s a great article titled, “We’re all Wittgensteinians now” that also goes into this debate of where meaning resides with specialist reference to LLMs like chatGPT.
       
 (DIR) Post #AU0WV8mYS7gQX8GA64 by budgibson@me.dm
       2023-03-26T12:29:57Z
       
       0 likes, 0 repeats
       
       @TedUnderwood People are also poor at tedium. That is potentially a clear win.
       
 (DIR) Post #AU0ZCkcFrYBmEXMfoW by budgibson@me.dm
       2023-03-26T13:00:14Z
       
       0 likes, 0 repeats
       
       @TedUnderwood LLMs are also really good at helping you fake it. The hot question being debated right now (and a bit the point of your post) is whether and when the faking is adequate. Faking is about adhering to a common pattern. Transcription and translation are two domains that seem to benefit tremendously from this property. They’re also tedious.
       
 (DIR) Post #AU2tQ2RZHIeSgltocC by budgibson@me.dm
       2023-03-27T14:26:23Z
       
       0 likes, 0 repeats
       
       @paul TBH, I think you’re getting out over your skis here. People will still have to develop competencies. It’s just that those competencies will be different to suit an AI-enhanced environment. What you teach and what people learn will be different.
       
 (DIR) Post #AU2tQ3ZP5eVuBMZaT2 by budgibson@me.dm
       2023-03-27T15:15:10Z
       
       0 likes, 0 repeats
       
       @paul I’m reminded of @simon’s recent post where he got chatGPT to crank out some AppleScript without himself knowing AppleScript, a win. That was one very specific test done by a highly skilled practitioner. I think a test of your assertion would be to take a naive user (we’ll have to define what that means) and have them do a similar task. What they have to know to do is what you have to teach them.
       
 (DIR) Post #AU9oz8cYiBiPXy7wMS by budgibson@me.dm
       2023-03-31T00:09:28Z
       
       0 likes, 0 repeats
       
       @TedUnderwood My instance requires content warnings for nudity. Things like that seem reasonable. Beyond that, I think you just unfollow if the person is producing stuff you don’t want to see. Filtering is another potential option. I’d like to see the ability to algorithmically filter my feed using something with chatGPT like capabilities. I’d
       
 (DIR) Post #AUU0DRMhf48dU1qWKu by budgibson@me.dm
       2023-04-09T17:47:39Z
       
       0 likes, 0 repeats
       
       @simon I think it was clear smartphones could take off at the time of blackberry. At the time, the clear smartphone value proposition was as an omni-communication device. I’m not sure anyone saw that smartphones would become the main information processing device for virtually everyone on the planet.
       
 (DIR) Post #AUZxhn0Sd8RoNMywEK by budgibson@me.dm
       2023-04-12T14:49:35Z
       
       0 likes, 0 repeats
       
       @grammargirl I just got my academic license for GitHub copilot. ChatGPT is very effective in some areas. The trick is to narrow its scope and not believe in it as gospel. Of course, my students and I have the advantage of knowing its fallible, and our code must pass the performance test.
       
 (DIR) Post #AUZyYqOGmMl6DePtT6 by budgibson@me.dm
       2023-04-12T14:59:13Z
       
       0 likes, 0 repeats
       
       @grammargirl A partial solution is that, as a user, you are responsible for how you use the tool. But, I agree, something that can basically slander people on the open web (e.g., saying public officials have been convicted of corruption when they have not) needs something to rein it in.
       
 (DIR) Post #AUd2FlJlrvOrBh7BpY by budgibson@me.dm
       2023-04-14T02:24:45Z
       
       0 likes, 0 repeats
       
       @TedUnderwood @Gargron Hard to see how that happens as bluesky is not particularly open. I do see room for instances to implement something like this.
       
 (DIR) Post #AUd2m1sJqROro0f1Lk by budgibson@me.dm
       2023-04-14T02:30:36Z
       
       0 likes, 0 repeats
       
       @TedUnderwood @Gargron FWIW, I think you’ll start to see some of this.
       
 (DIR) Post #AUkcBLtD5PDzewKOem by budgibson@me.dm
       2023-04-17T15:52:02Z
       
       0 likes, 0 repeats
       
       @grammargirl Are you familiar? She’s on NPR (1A) today. Professor | Valerie Fridland, PhD https://www.valeriefridland.com/
       
 (DIR) Post #AUkcBMP7AmkHFtFsO0 by budgibson@me.dm
       2023-04-17T17:33:58Z
       
       0 likes, 0 repeats
       
       @grammargirl Am now looking on her website with a more thorough eye and see your review of her book and realize you are, indeed, familiar.
       
 (DIR) Post #AUkfV1Ao7Mip3zczrs by budgibson@me.dm
       2023-04-17T18:47:25Z
       
       0 likes, 0 repeats
       
       @grammargirl I’ll be sure to give that one a listen. She was great on 1A. Am going to get the book. Her discussion of how English came to be a formalized language was something else. At least we don’t have an Académie Française style institution.
       
 (DIR) Post #AVIGlKhrN13onXPNhI by budgibson@me.dm
       2023-05-03T23:50:27Z
       
       0 likes, 0 repeats
       
       @grammargirl Really, either way. It’s a power relationship. Who has more need of the other? That’s not always a question of seniority.
       
 (DIR) Post #AWqhKfYc8sAa9UOV28 by budgibson@me.dm
       2023-06-19T12:20:26Z
       
       0 likes, 0 repeats
       
       @grammargirl I was just listening to your podcast on the similarity of the words for father across languages where you mentioned Grimm’s Law in the p -> f and t -> th transitions in Indo-European languages. Probably, the most interesting observation though was how languages have front of mouth sounds for dad (and mom too).